US20160072472A1 - Method and device for bi-state control of nonlinear resonators - Google Patents

Method and device for bi-state control of nonlinear resonators Download PDF

Info

Publication number
US20160072472A1
US20160072472A1 US14/627,801 US201514627801A US2016072472A1 US 20160072472 A1 US20160072472 A1 US 20160072472A1 US 201514627801 A US201514627801 A US 201514627801A US 2016072472 A1 US2016072472 A1 US 2016072472A1
Authority
US
United States
Prior art keywords
signal
resonator
voltage
drive
frequency
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
US14/627,801
Other versions
US10608614B2 (en
Inventor
Gary K. Fedder
Congzhong Guo
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Carnegie Mellon University
Original Assignee
Carnegie Mellon University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Carnegie Mellon University filed Critical Carnegie Mellon University
Priority to US14/627,801 priority Critical patent/US10608614B2/en
Assigned to CARNEGIE MELLON UNIVERSITY, A PENNSYLVANIA NON-PROFIT CORPORATION reassignment CARNEGIE MELLON UNIVERSITY, A PENNSYLVANIA NON-PROFIT CORPORATION ASSIGNMENT OF ASSIGNORS INTEREST (SEE DOCUMENT FOR DETAILS). Assignors: GUO, CONGZHONG, FEDDER, GARY K.
Publication of US20160072472A1 publication Critical patent/US20160072472A1/en
Assigned to NATIONAL SCIENCE FOUNDATION reassignment NATIONAL SCIENCE FOUNDATION CONFIRMATORY LICENSE (SEE DOCUMENT FOR DETAILS). Assignors: CARNEGIE-MELLON UNIVERSITY
Application granted granted Critical
Publication of US10608614B2 publication Critical patent/US10608614B2/en
Active legal-status Critical Current
Adjusted expiration legal-status Critical

Links

Images

Classifications

    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03HIMPEDANCE NETWORKS, e.g. RESONANT CIRCUITS; RESONATORS
    • H03H9/00Networks comprising electromechanical or electro-acoustic devices; Electromechanical resonators
    • H03H9/02Details
    • H03H9/02244Details of microelectro-mechanical resonators
    • H03H9/02259Driving or detection means
    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03JTUNING RESONANT CIRCUITS; SELECTING RESONANT CIRCUITS
    • H03J1/00Details of adjusting, driving, indicating, or mechanical control arrangements for resonant circuits in general
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01LMEASURING FORCE, STRESS, TORQUE, WORK, MECHANICAL POWER, MECHANICAL EFFICIENCY, OR FLUID PRESSURE
    • G01L1/00Measuring force or stress, in general
    • G01L1/10Measuring force or stress, in general by measuring variations of frequency of stressed vibrating elements, e.g. of stressed strings
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N29/00Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
    • G01N29/02Analysing fluids
    • G01N29/022Fluid sensors based on microsensors, e.g. quartz crystal-microbalance [QCM], surface acoustic wave [SAW] devices, tuning forks, cantilevers, flexural plate wave [FPW] devices
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N29/00Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
    • G01N29/02Analysing fluids
    • G01N29/036Analysing fluids by measuring frequency or resonance of acoustic waves
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N29/00Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
    • G01N29/04Analysing solids
    • G01N29/12Analysing solids by measuring frequency or resonance of acoustic waves
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01CMEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C19/00Gyroscopes; Turn-sensitive devices using vibrating masses; Turn-sensitive devices without moving masses; Measuring angular rate using gyroscopic effects
    • G01C19/56Turn-sensitive devices using vibrating masses, e.g. vibratory angular rate sensors based on Coriolis forces
    • G01C19/5776Signal processing not specific to any of the devices covered by groups G01C19/5607 - G01C19/5719
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2291/00Indexing codes associated with group G01N29/00
    • G01N2291/02Indexing codes associated with the analysed material
    • G01N2291/024Mixtures
    • G01N2291/02491Materials with nonlinear acoustic properties
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01NINVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N2291/00Indexing codes associated with group G01N29/00
    • G01N2291/02Indexing codes associated with the analysed material
    • G01N2291/025Change of phase or condition
    • G01N2291/0256Adsorption, desorption, surface mass change, e.g. on biosensors

Definitions

  • the field of the invention relates generally to non-linear resonators.
  • Resonant-based sensing has long been exploited to gauge micromechanical system parameters such as mass and spring constant. Gravimetric sensors and mass flow meters that measure the mass change and the stress sensors that link to suspension spring constant changes all exemplify this approach.
  • harmonic resonators suffer from a direct link between amplitude and frequency noise that degrades sensitivity
  • the harmonic and parametric resonators differ in the way the resonance is excited.
  • the harmonic resonator has a fixed spring constant, with an external drive that is sinusoidal in time t and invariant in displacement x.
  • the parametric resonator is driven by a nonlinear electrostatic drive at an electrical frequency that creates a displacement-dependent spring constant k eff .
  • Parametric resonance conditions are satisfied when the periodic variation of k eff is at twice the natural mechanical resonance frequency or at other integer multiples fraction of twice the natural mechanical resonance frequency (i.e., 2 ⁇ r /n, where ⁇ r is the natural mechanical resonant frequency and n is an integer larger or equal to 1).
  • a microelectromechanical parametric resonator operates in two states—one “off” state that has zero motion and the other “on” state whose amplitude increases exponentially over time. In the “off” state, the resonator is nominally perfectly balanced at rest. However, because of the existence of thermomechanical noise or with the existence of external disturbances, a very small but non-zero noise displacement typically on the order of femtometers to nanometers is present, depending on the level of noise inherent in the system.
  • the mass-spring-damper system With non-zero displacement and application of a parametric drive voltage under appropriate pump frequencies where the parametric resonance condition is satisfied, the mass-spring-damper system periodically increases (hardens) or decreases (softens) its effective spring force, resulting in the mass being pumped into parametric resonance.
  • a drive scheme is needed to excite the parametric resonance and control the servo such that the controlled setpoint along the bifurcation “jump” is maintained.
  • Bifurcation-based control approach can be realized both open-loop and closed-loop. The former is slow, sometimes taking up to minutes since the parametric drive frequency is swept with fine increments until a jump in amplitude is observed, and is impractical for most practical (real-time) sensing applications.
  • the closed-loop scheme in principle can servo along the jump point (for example, the point along A-B in FIG. 1 ) enabling rapid readout of the bifurcation amplitude and/or frequency.
  • One analog bifurcation control approach takes the phase variance as the input of the controller. These controllers are generally complex and difficult for on-chip implementation. It also produces a relatively small amplitude on the order of 100 pm.
  • the advantage of the controller of the present invention is its practical feasibility for on-chip circuit implementation and its ability to servo at relatively large amplitude (typically about four orders of magnitude larger in the displacement amplitude than the 100 pm amplitude) along the sharp jump of the bifurcation without altering the plant.
  • a bi-state bifurcation-based control system comprising a resonator 100 and a controller 301 is described that is useful for ultrasensitive micro-electromechanical systems (MEMS) applications and addresses issues of hysteresis and long settling time.
  • MEMS micro-electromechanical systems
  • One key aspect of this invention is the use of a bi-state switching that is set to a speed much faster than the amplitude response dynamics of the resonator 100 . Switching is accomplished by modulating the spring constant of the parametric resonator 100 using two distinct drive voltage levels and a pulse width modulator 302 to control alternating between each drive voltage level.
  • the servo setpoint 500 is chosen along the sharp jump transition, i.e., along the line between A and B or between S and P as indicated in FIG. 1 .
  • the two dots are two possible setpoints 500 ; a multitude of different setpoints 500 along the bifurcation lines A-B and S-P could be chosen. However, only one setpoint 500 is controlled by a corresponding bi-state control system at a given time.
  • a parametric resonator 100 is driven by a DC parametric drive voltage, V dc , in series with an AC excitation voltage, V ac .
  • V dc DC parametric drive voltage
  • V ac AC excitation voltage
  • the principle of this control technique is to fix the frequency and amplitude of V ac such that the system is driven into either the “on” or “off” state with the appropriate corresponding value of V dc .
  • V ac is a sinusoidal signal. However the signal could be a square wave, a triangle wave, or any other periodic waveform that has a fundamental component at the appropriate frequency.
  • the DC bias voltage is controlled by the feedback and constrained to take one of the two values in a pulse modulation scheme.
  • the pulse width modulation voltage, ⁇ PWM effectively takes on the role of V dc but having two distinct DC values.
  • the constraint that the parametric drive amplitude ⁇ PWM is only allowed to take one of the two possible values (one in the “on” region, and the other in the “off” region) highlights the key feature of this control scheme.
  • the feedback pulse frequency is fixed and the duty cycle is modulated (i.e., pulse width modulation is employed).
  • pulse width modulation is employed.
  • this approach also achieves a high ratio of resonant amplitude to the minimum detectable amplitude signal with much higher controlled amplitude.
  • control system states for a Duffing microresonator 100 are set through bi-state amplitude modulation of the AC drive gain V ac while operating at a fixed drive frequency such that one of the system states is within an “instability tongue” whereas the second state resides outside of this tongue, shown in FIG. 5 .
  • the tongue draws the boundary between the high displacement and low displacement regions.
  • the DC drive voltage and the frequency are fixed.
  • the amplitude servo point B is chosen at the maximum “on” amplitude prior to the bifurcation jump event.
  • FIG. 1 is a chart illustrating steady-state frequency response of a nonlinear parametric resonator.
  • FIG. 2 is the block diagram of the of the bi-state control system, according to one embodiment, with a Silicon-On-Insulator (SOI) MEMS parametric resonator as the plant, together with scanning electron microscope (SEM) views of the parametric drive fingers (D) and capacitive sense fingers (S) (shown as insets).
  • SOI Silicon-On-Insulator
  • SEM scanning electron microscope
  • FIG. 3 is an optical microscope image of a microresonator with zoomed-in SEM views of a drive comb and a crab-leg spring.
  • FIG. 4 is a chart illustrating hysteresis of a Duffing resonator, indicating point A as the down-sweep bifurcation frequency and point B as a control point that is just below the up-sweep bifurcation frequency.
  • FIG. 5 is a bifurcation diagram of the Duffing resonator showing the locus of the up and down bifurcation points, and the states for the control, where one is located in the high displacement region and the other in the low displacement region.
  • FIG. 6 is the block diagram of a bi-state Duffing resonator control system according to one embodiment.
  • FIG. 7 depicts charts of the turn-on transient of the bi-state control loop for a Duffing resonator.
  • FIG. 8 depicts charts showing the recapture of the Duffing resonance by rapidly switching back to the “on” state.
  • FIG. 9 depicts charts illustrating the turn-off of the Duffing resonance by increasing the off-state time beyond the critical off-state time.
  • FIG. 10 is the zoom-in view of the frequency down sweep of a parametric resonator in the time domain, where the system state has to travel through the slow time manifold from “off” state (point A) to reach the “on” state (point B).
  • FIG. 11A illustrates the measured bifurcation diagram of the “on” (shaded) and “off” regions for a parametric resonator. Callout shows two servo states that are marked as circles.
  • FIG. 11B is a composite view that shows the relationship between the measured bifurcation of a parametric resonator and the corresponding frequency swept resonant amplitudes and that indicates the location of possible operating points at a select drive frequency.
  • FIG. 12 is a series of charts demonstrating the recapturing of the parametric resonance “zero-vicinity” state.
  • FIG. 13 illustrates the turn-on transient of the bi-state control loop for the parametric control embodiment.
  • FIG. 14 illustrates shifts in the servo setpoint for a parametric resonator illustrating closed-loop dynamics. Solid curves are obtained with only proportional control and the dash curves are obtained using proportional-derivative control.
  • FIG. 15 represents an Allan deviation plot of the temperature compensated controller duty cycle output quantifying the stability of a particular embodiment.
  • FIG. 16 is a flow diagram of the method of one embodiment of the present invention.
  • a control system and method of the present invention makes possible the highly sensitive detection of mass or strain change by utilizing the bifurcation behavior inherent in a class of nonlinear parametric resonators 100 and in Duffing resonators 100 .
  • the bifurcation behavior is represented as sharp increase in amplitude during a frequency sweep.
  • the bifurcation jump 501 occurs along the lower frequency transition or higher frequency transition.
  • FIG. 2 illustrates a block diagram of the bi-state bifurcation control system to serve as a general description of the resonator 100 and control loop 301 and is described in detail in the embodiments below.
  • the system is comprised of a parametric resonator 100 having capacitive drive combs 104 and capacitive sense combs 105 .
  • a direct current (DC) drive voltage is applied across the drive combs 104 to set the effective spring constant of the resonator 100 .
  • the sense combs 105 provide information used in a control loop 301 that switches the resonator 100 between “on” and “off” states.
  • the combs 104 and 105 are comprised of a primary set of comb fingers 109 attached to the mass 102 and a corresponding set of fingers 108 that are mechanically anchored.
  • the system further comprises a control loop 301 , which may include a drive circuit 310 .
  • the controller, or control loop, 301 further comprises a feedback circuit 308 and an error circuit 309 .
  • the method of control of the resonator 100 is: at step 800 , sensing a differential motional output current from the sense capacitors 105 of a MEMS parametric resonator 100 ; at step 801 , generating a feedback signal; at step 802 , extracting displacement and phase signal based on the resonator 100 output (in this embodiment using a lock-in amplifier 303 with a low-pass filter 307 (LPF) having a 2 Hz cutoff frequency); at step 803 , subtracting the displacement signal, ⁇ x , with a pre-set threshold, V thresh , and generating an error signal, ⁇ amp ; at step 804 , inputting the error signal into the proportional-integral-derivative controller; at step 805 , generating the control signal with the proportional-integral-differential (PID) controller 306 based on the error signal; at step 806 , inputting the control signal into the pulse-width-
  • PID proportional-integral-different
  • a parametric control scheme is used to control a MEMS resonator 100 operating under nonlinear parametric resonance conditions.
  • the parametric control scheme addresses the issue of the slow time manifold by servoing along this manifold with a bi-state PWM switching that is set to a speed much faster than the amplitude response dynamics of the resonator 100 .
  • the PWM switching speed is set to 200 kHz.
  • other switching speeds can be used as long as the speed is faster than the resonator 100 system dynamics.
  • the servo point 500 is chosen to take a non-zero displacement amplitude (1.9 ⁇ m, shown as point C in FIG.
  • the servo point 500 lies along the bifurcation jump 501 A-B in FIG. 1 .
  • the servo point 500 can be set along the bifurcation jump point 501 S-P in FIG. 1 , in an alternate embodiment.
  • control plant 101 is a parametric resonator 100 made in a Silicon-On-Insulator (SOI)-MEMS fabrication process with a 15 ⁇ m-thick structural silicon device layer forming a micromechanical mass-spring system.
  • SOI Silicon-On-Insulator
  • Two primary sets of electrostatic comb fingers 109 are attached to the plate mass 102 .
  • Corresponding sets of comb fingers 108 are interdigitated with the primary set of comb fingers 109 attached to the mass 102 to form a drive comb 104 and a sense comb 105 .
  • the electrostatic sense fingers 109 which forms part of the sense comb 105 (a scanning electron microscope (SEM) image shown in FIG.
  • inset S have straight sidewalls and act as capacitors whose capacitance varies linearly with displacement.
  • the electrostatic drive comb fingers 109 which form part of the drive comb 104 (a SEM image shown in FIG. 2 , inset D) is specially shaped along its sidewalls (for example, non-straight sidewalls) to create a voltage controlled linear spring constant ⁇ 1 in the force, F p , acting on the resonator 100 in Eq. (1) that is necessary to excite the parametric resonance.
  • a cubic-term spring constant ⁇ 3 in Eq. (1) provides specific nonlinear behavior that limits the parametric resonance amplitude.
  • the specific effective system nonlinearity can be either hardening (i.e., ⁇ 3 >0) or softening (i.e., ⁇ 3 ⁇ 0), which is determined by the combined mechanical and electrostatic nonlinearity.
  • the exemplary embodiment of the control plant 102 in FIG. 2 has a hardening nonlinearity.
  • the bi-state control principle applies to both hardening and softening nonlinear resonators.
  • a DC bias voltage V dc,set is applied to a total of 278 moving straight fingers 109 that are attached to a mass 102 of 9 ⁇ g.
  • the DC bias voltage sets the effective spring constant of the resonator 100 .
  • a voltage ⁇ shape formed by summing the DC voltage from the PWM 302 , ⁇ PWM , with an AC excitation voltage amplitude, V ac is applied on the shaped fingers 108 that are mechanically anchored.
  • the other side of the comb 104 i.e.
  • TIA transimpedance amplifiers 304
  • the DC parametric polarization voltage V dc,total takes two values, one value that drives the device into parametric resonance, and the other value drives the resonator to zero displacement.
  • the bi-state control is accomplished by fixing the parametric drive frequency and the AC drive amplitude V ac , such that the system can be readily moved between “on” and “off” states with appropriate value of V dc,total (circles in FIG. 11B ).
  • the constraint that the parametric polarization voltage is only allowed to take one of the two possible values (one in the “on” region, and the other in the “off” region) highlights a key characteristic of this control scheme.
  • a linear control scheme having a continuum of polarization voltage levels between the two state values would result in an unstable controller and would not servo on the desired setpoint. Instead, such a control system would latch into one of the “latched”, steady-state on (s.s. on in FIG. 11B ) or steady-state off (s.s. off in FIG. 11B ) positions and thus would not track the bifurcation as an average of the PWM 302 output.
  • a first DC parametric polarization voltage V dc,total 49.7 V lands inside the “on” region, starting to move the state along the slow time manifold in FIG. 10 from point A to B.
  • a second DC parametric polarization voltage V dc,total 53.2 V is applied which resides in the “off” region, moving the state back from point B to A.
  • This “recapture” dynamics is shown in FIG. 12 .
  • the successful recapturing of the resonator 100 state is guaranteed provided that the “on-state” time F is kept shorter than 602.8 ms, in this particular example. The recapturing breaks down if the on-state time is set beyond this critical time.
  • recapturing occurs by rapidly switching back to the “off” state through the use of the control loop 301 , with the pulse width modulator 302 alternating the DC voltage based on a set clock speed, as illustrated in FIG. 2 .
  • a shaped-finger comb parametric resonator 100 serves as the control plant 101 .
  • this twice resonance frequency component drives the oscillator into parametric resonance.
  • the amplitude and the phase of the velocity signal ⁇ x picked off by the TIA 304 , are measured by a lock-in amplifier (LIA) 303 which utilizes the signal at ⁇ r to demodulate the oscillation velocity signal at resonance.
  • LIA lock-in amplifier
  • This error signal is fed into a proportional-integral-derivative (PID) controller 306 that provides the duty cycle information to the pulse-width-modulation (PWM) module 302 .
  • PWM 302 generates a square waveform with a fixed frequency at 200 kHz, outputting either high or low voltages that correspond to the first DC voltage or the second DC voltage.
  • the PID 306 coefficients are chosen in such a way that the width ratio of the “on” and “off” states is changed to maintain the servo point between the maximum “on” and zero “off” states.
  • This example shows the servo 500 at point C in FIG. 10 with 1.9 ⁇ m amplitude to yield a large ratio of servo amplitude to the minimum detectable amplitude.
  • the overall parametric drive voltage across the drive comb fingers 104 is set by the PWM signal ⁇ PWM , the AC drive ⁇ ac , and the DC bias voltage V dc on the mass 102 .
  • ⁇ parametric
  • the duty cycle signal is an ultra-sensitive indicator of the servo setpoint 500 along the slow sinusoidal-amplitude dynamics, therefore it is chosen as the controller output ⁇ o , derived by sending the PWM signal through a low pass filter (LPF) 307 .
  • LPF low pass filter
  • FIG. 13 Turn-on dynamics of this bi-state control loop in a vacuum of 18 mTorr, by way of one particular example, is shown in FIG. 13 .
  • a step change in the parametric drive frequency turns the resonator 100 to the “on” state (2.2 ⁇ m) from the “off” state.
  • the controller 301 is turned on at 23.3 s by enabling the PID controller 306 .
  • the servo point 500 (1.9 ⁇ m) backs off from the maximum “on” and settles on the slow time manifold.
  • the duty cycle rings between 0.69 to 0.2 as the servo turns on and then settles to a steady-state control value of around 0.34.
  • Closed-loop dynamics shown in FIG. 14 illustrate the effect of the derivative controller coefficient on the closed-loop dynamics.
  • the servo threshold, V thresh is changed at 0 s from 25 mV (equivalent to 1.75 ⁇ m in displacement amplitude
  • V thresh The servo threshold
  • the displacement amplitude changes from 1.89 ⁇ m to 2.04 ⁇ m with a significant amount of ringing (solid line).
  • the bi-state control scheme for parametric resonators 100 is general and applies to any bi-state feedback scheme, of which the highlighted PWM scheme is a good example.
  • Other possible feedback schemes include pulse-density modulation, for example in delta modulator or delta-sigma modulator architectures.
  • pulse-density modulation for example in delta modulator or delta-sigma modulator architectures.
  • Other discrete multi-state control schemes could also be employed instead of simply two states in feedback.
  • An example would be in pulse amplitude modulation with a finite number of states.
  • the preceding list is not meant to be exhaustive as many other digital and discrete-amplitude feedback methods exist.
  • the system is a bi-state control system capable of servoing on the falling edge of the microelectromechanical Duffing resonant bifurcation setpoint.
  • the system is presented with a 15 ⁇ m-thick SOI-MEMS Duffing resonator 100 as the control plant 101 , which is comprised of a shuttle mass 102 and a set of capacitive comb fingers 109 .
  • FIG. 3 is a perspective view of the lateral capacitive combs 104 and 105 electrode geometry of the control plant 101 .
  • the capacitive combs 104 and 105 are symmetrically placed on both sides of the shuttle mass 102 ; the left set serves as the drive electrode 104 and the right set serves as the capacitive pick-off sense electrode 105 .
  • the perforated shuttle mass 102 is suspended by four symmetric crab-leg springs, or spring suspension, 106 .
  • the nonlinearity inherent in the system stems from the suspension and the nonlinear electrostatic spring constants arising from the comb drive geometry.
  • the typical measured frequency response of a Duffing resonator 100 is plotted in FIG. 4 .
  • symbolizes the frequency up-sweep and ⁇ is the frequency down-sweep.
  • This class of nonlinear resonator 100 exhibits a hysteresis effect when the frequency is swept bi-directionally.
  • the numbers depicted in FIG. 4 correspond to a test performed in a low-pressure environment of 25 mTorr to serve as an illustrative example.
  • the resonance peak bends toward higher frequencies, yielding multi-valued solutions for any particular frequency value located around the resonance. For example, on the edge of the bent peak at the maximum amplitude (point B on FIG. 4 ), a jump in the resonance amplitude from 4.8 ⁇ m to 0 ⁇ m occurs if the drive frequency is increased an infinitesimal amount.
  • the frequency at which the instantaneous jump occurs is called the “bifurcation frequency” f B , which equals 4167 Hz for the setpoint B in FIG. 4 .
  • the bifurcation frequency can vary depending on the design of the resonator.
  • the jump 501 happens at a lower frequency from 0 ⁇ m to 1.0 ⁇ m (point A), indicative of a hysteresis effect.
  • the region I between points A and B defines the region where a high amplitude response occurs in response to a frequency change.
  • region II To the right of point B (region II), the resonance amplitude reduces to very low displacement.
  • the states for a bi-state control loop can be defined by adjusting the AC drive voltage V ac which shifts the entire frequency response curve in FIG. 4 , thus resulting in changes in the bifurcation points A and B.
  • V ac AC drive voltage
  • FIG. 5 which defines the “instability tongue” with the locus of the amplitude jump points.
  • the right curve is the up-sweep bifurcation frequency (analogous to point B in FIG. 4 ) showing the falling edge of the instability; the left curve is the down-sweep bifurcation frequency (point A).
  • the DC polarization voltage is fixed at 10 V in the example loci in FIG. 5 .
  • the changes in the drive voltages modify the system's effective spring constant, which affects the bending of the resonance curve and the regions where the system enters instability.
  • the drive voltage shapes the instability tongue.
  • the tongue width measured in units of frequency (e.g., Hertz) depends on the AC drive amplitude.
  • the position of the tongue along the frequency axis is a strong function of the DC polarization voltage.
  • the points 502 in FIG. 5 identify the states for a bi-state control loop.
  • One of the states (“on” state) is inside the “tongue” that has a large amplitude response; the other state (“off” state) resides outside the “tongue” representing a low amplitude response.
  • the two states are on opposite sides of the bifurcation point in steady-state operation. Adjusting the AC voltage, V ac , near the boundary can move the control state between high and low displacement values. The rapid switching between these two states determines the servo setpoint.
  • the control loop 301 is schematically shown in FIG. 6 .
  • the control setpoint 500 of the loop 301 is chosen at point B in FIG. 4 .
  • the AC drive frequency and the DC bias are fixed to values such that the system can be moved between high and low displacement regions with appropriate bi-state values of AC amplitude
  • the Duffing resonator 100 is driven by V ac whose amplitude and frequency are set by the pulse-width-modulator (PWM) 302 output and the lock-in amplifier (LIA) 303 , respectively.
  • PWM pulse-width-modulator
  • LIA lock-in amplifier
  • a transimpedance amplifier 304 followed by the LIA 303 picks up the resonator velocity amplitude
  • the loop 301 tracks
  • the error output is fed into a PID (proportional-integral-derivative) controller 306 to generate the input signal to the PWM 302 representing the pulse-train duty cycle.
  • the output of the PWM 302 generates the bi-state control signal that sets the AC drive amplitude.
  • the V ac is controlled by the duty-cycle feedback and rapidly adjusts the oscillation between the two states (e.g., high and low displacements) at an example PWM rate of 300 kHz to achieve the servo at the maximum displacement value without turning off the resonator 100 .
  • This resettling time is inversely proportional to the damping and is typically on the order of 1 s for the resonator 100 operated under a similar vacuum level.
  • FIG. 7 shows a series of graphs which depict the turn-on transient of the closed-loop controller 301 with an initial approaching phase.
  • the drive AC voltage is turned on at 15 s, followed by an approach phase when the frequency is increased from 4123 Hz to 4167 Hz.
  • the controller 301 is turned on afterwards.
  • the controller 301 is turned off by switching off V ac .
  • region (1) the resonator is off.
  • region (2) a sinusoidal voltage at 4123 Hz is applied to the resonator 100 drive fingers of comb 104 , resulting in a step turn on of the displacement amplitude to a non-zero value due to the capacitive feedthrough.
  • Region (3) is the necessary “approach” stage when the resonator 100 is turned on to the maximum displacement operation point (point B in FIG. 4 ) by increasing the drive frequency from 4123 Hz to 4167 Hz.
  • the PID controller 306 threshold V thresh 305 is set in region (4) after reaching this maximum operation point, turning on the servo, after which time the PWM 302 output rapidly changes between 0.3 V and 3.4 V, with duty cycle centered on 0.5.
  • the servo is switched off by turning off V ac .
  • the phase is well-defined during the time the controller 301 is on and loses track of the input V ac when the controller 301 is turned off.
  • a detailed analysis of the dynamics on the edge of the bifurcation jump 501 provides information on how fast the controller 301 should switch between two states to guarantee successful recapturing of the “on” state. These “off-state” dynamics are characterized in FIG. 8 and FIG. 9 .
  • V ac In open-loop operation, a step change in V ac is introduced from 25.6 mV to 24.1 mV, at which voltage the resonator 100 is not fully in the off state but rather just starting to turn off. V ac is then turned back to 25.6 mV after a prescribed amount of time ⁇ .
  • is less than 394.6 ms, the Duffing resonance is successfully recaptured as presented in FIG. 8 .
  • When ⁇ is larger than 395.0 ms, the resonance falls back to the low displacement state.
  • the critical off-state time ⁇ at which the resonance is on the edge of being recaptured is 394.6 ms.
  • the critical off-state time is 394.6 ms.
  • the turn-off time is 395.0 ms.
  • Allan standard deviation of the controller 301 duty cycle output for a parametric resonator 100 is shown in FIG. 15 that represents the servo point 500 variation versus the integration time, ⁇ , and demonstrates the stability of the controller 301 .
  • the duty cycle signal is temperature compensated to eliminate the effect of temperature on the bifurcation frequency variation and hence the servo setpoint 500 .
  • the data shown in FIG. 15 represent a test performed in a constant vacuum of 18 mTorr.
  • the minimum detectable duty cycle found at an integration time of 4.3 s is 2.36 ⁇ 10 ⁇ 4 , which corresponds to 4.48 ⁇ resolution in displacement amplitude
  • the large displacement (1.9 ⁇ m) operation yields a much higher ratio of operation amplitude to the minimum detectable signal compared with recent bifurcation-based control schemes.
  • the embodiments and examples described herein demonstrate a nonlinear control of parametric and Duffing resonators 100 using a bi-state switching methodology, and further builds an ultra-sensitive platform for detecting parametric parameter changes that are indicative of resonance frequency shifts.
  • the advantage of the system and method of the present invention is its design allowing for on-chip or off-chip implementation, and its ability to servo at relatively large amplitude (roughly four orders of magnitude larger in the displacement amplitude) along the sharp jump of the bifurcation without altering the plant.
  • This bi-state closed-loop control system can be used in any application where mechanical resonant sensors are used now.

Abstract

A bi-state bifurcation-based control system and method for nonlinear resonators, which utilizes a control loop to servo on the edge of the bifurcation jump, either at the maximum “on” point prior to the Duffing bifurcation jump or along the rising edge of the parametric bifurcation.

Description

    CROSS-REFERENCE TO RELATED APPLICATIONS
  • This application claims the benefit under 35 U.S.C. §119 of U.S. Provisional Application Ser. No. 61/966,280, filed Feb. 20, 2014, which is incorporated by reference herein in its entirety.
  • STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH
  • This invention was made with government support under the National Science Foundation CNS 0941497. The government has certain rights in this invention.
  • BACKGROUND OF THE INVENTION
  • The field of the invention relates generally to non-linear resonators. Resonant-based sensing has long been exploited to gauge micromechanical system parameters such as mass and spring constant. Gravimetric sensors and mass flow meters that measure the mass change and the stress sensors that link to suspension spring constant changes all exemplify this approach. As harmonic resonators suffer from a direct link between amplitude and frequency noise that degrades sensitivity, it would be advantageous to use a class of nonlinear parametric resonators for developing an extremely sensitive mechanism to reveal tiny changes in resonance characteristics. This improved sensitivity stems from an instantaneous “jump” transition from near-zero to high displacement in parametric resonators, yielding high ratio of resonant amplitude to the minimum detectable amplitude signal.
  • The harmonic and parametric resonators differ in the way the resonance is excited. The harmonic resonator has a fixed spring constant, with an external drive that is sinusoidal in time t and invariant in displacement x. The parametric resonator is driven by a nonlinear electrostatic drive at an electrical frequency that creates a displacement-dependent spring constant keff. Parametric resonance conditions are satisfied when the periodic variation of keff is at twice the natural mechanical resonance frequency or at other integer multiples fraction of twice the natural mechanical resonance frequency (i.e., 2ωr/n, where ωr is the natural mechanical resonant frequency and n is an integer larger or equal to 1).
  • A microelectromechanical parametric resonator operates in two states—one “off” state that has zero motion and the other “on” state whose amplitude increases exponentially over time. In the “off” state, the resonator is nominally perfectly balanced at rest. However, because of the existence of thermomechanical noise or with the existence of external disturbances, a very small but non-zero noise displacement typically on the order of femtometers to nanometers is present, depending on the level of noise inherent in the system. With non-zero displacement and application of a parametric drive voltage under appropriate pump frequencies where the parametric resonance condition is satisfied, the mass-spring-damper system periodically increases (hardens) or decreases (softens) its effective spring force, resulting in the mass being pumped into parametric resonance.
  • An instantaneous bifurcation “jump” is seen from a near-zero “off” state (point A) to a non-zero high amplitude (point B) “on” state in the representative steady-state amplitude versus frequency characteristic in FIG. 1. This sharp transition from A to B is called “bifurcation”. Quality factor (Q), which is inversely related to energy loss per cycle, is an important metric used to evaluate the performance of a resonator. Lower loss generally gives rise to a sharper amplitude peak in the frequency spectrum. The infinitely steep slope on the bifurcation edge effectively enhances Q-factor. This intriguing property promotes many ultra-sensitive sensor applications such as mass flow rate sensors, gravimetric sensors, magnetometers and strain sensors, and is used in noise-squeezing applications for parametric amplification.
  • An analogous bifurcation response occurs in a micromechanical Duffing resonator, where the bending of the resonance amplitude curve is due to the mechanical nonlinearity. The bending leads to multi-valued solutions for one particular frequency, which results in “bi-stability” in frequency response.
  • To enable parametric resonance and/or Duffing resonance to be used in making an ultra-sensitive mass sensor or strain gauge, a drive scheme is needed to excite the parametric resonance and control the servo such that the controlled setpoint along the bifurcation “jump” is maintained. Bifurcation-based control approach can be realized both open-loop and closed-loop. The former is slow, sometimes taking up to minutes since the parametric drive frequency is swept with fine increments until a jump in amplitude is observed, and is impractical for most practical (real-time) sensing applications. The closed-loop scheme in principle can servo along the jump point (for example, the point along A-B in FIG. 1) enabling rapid readout of the bifurcation amplitude and/or frequency. There are two practical limits to this approach. One is due to the hysteresis (jumps can occur at A-B or S-P depending on the direction of the frequency sweep), which inhibits the amplitude from readily moving back to its original state once the amplitude has transitioned across the bifurcation. The other limitation originates from a slow-time manifold that the resonator state has to travel in order to transit between “on” and “off” states. This slow-time manifold on the system trajectory is relatively long (on the order of seconds), inhibiting stable servo operation when linear feedback control approaches are applied.
  • One analog bifurcation control approach takes the phase variance as the input of the controller. These controllers are generally complex and difficult for on-chip implementation. It also produces a relatively small amplitude on the order of 100 pm. The advantage of the controller of the present invention is its practical feasibility for on-chip circuit implementation and its ability to servo at relatively large amplitude (typically about four orders of magnitude larger in the displacement amplitude than the 100 pm amplitude) along the sharp jump of the bifurcation without altering the plant.
  • BRIEF SUMMARY OF THE INVENTION
  • According to one embodiment of the present invention, a bi-state bifurcation-based control system comprising a resonator 100 and a controller 301 is described that is useful for ultrasensitive micro-electromechanical systems (MEMS) applications and addresses issues of hysteresis and long settling time. One key aspect of this invention is the use of a bi-state switching that is set to a speed much faster than the amplitude response dynamics of the resonator 100. Switching is accomplished by modulating the spring constant of the parametric resonator 100 using two distinct drive voltage levels and a pulse width modulator 302 to control alternating between each drive voltage level. The servo setpoint 500 is chosen along the sharp jump transition, i.e., along the line between A and B or between S and P as indicated in FIG. 1. The two dots are two possible setpoints 500; a multitude of different setpoints 500 along the bifurcation lines A-B and S-P could be chosen. However, only one setpoint 500 is controlled by a corresponding bi-state control system at a given time.
  • In one embodiment, a parametric resonator 100 is driven by a DC parametric drive voltage, Vdc, in series with an AC excitation voltage, Vac. The principle of this control technique is to fix the frequency and amplitude of Vac such that the system is driven into either the “on” or “off” state with the appropriate corresponding value of Vdc. In this embodiment, Vac is a sinusoidal signal. However the signal could be a square wave, a triangle wave, or any other periodic waveform that has a fundamental component at the appropriate frequency. The DC bias voltage is controlled by the feedback and constrained to take one of the two values in a pulse modulation scheme. In this case, the pulse width modulation voltage, υPWM, effectively takes on the role of Vdc but having two distinct DC values. The constraint that the parametric drive amplitude υPWM is only allowed to take one of the two possible values (one in the “on” region, and the other in the “off” region) highlights the key feature of this control scheme. In the embodiment shown in FIG. 2, the feedback pulse frequency is fixed and the duty cycle is modulated (i.e., pulse width modulation is employed). In addition to addressing the issues of limit cycling and instability, this approach also achieves a high ratio of resonant amplitude to the minimum detectable amplitude signal with much higher controlled amplitude.
  • In an alternative embodiment, shown in FIG. 6, the control system states for a Duffing microresonator 100 are set through bi-state amplitude modulation of the AC drive gain Vac while operating at a fixed drive frequency such that one of the system states is within an “instability tongue” whereas the second state resides outside of this tongue, shown in FIG. 5. The tongue draws the boundary between the high displacement and low displacement regions. The DC drive voltage and the frequency are fixed. In the representative example shown in FIG. 4, the amplitude servo point B is chosen at the maximum “on” amplitude prior to the bifurcation jump event.
  • BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS
  • FIG. 1 is a chart illustrating steady-state frequency response of a nonlinear parametric resonator.
  • FIG. 2 is the block diagram of the of the bi-state control system, according to one embodiment, with a Silicon-On-Insulator (SOI) MEMS parametric resonator as the plant, together with scanning electron microscope (SEM) views of the parametric drive fingers (D) and capacitive sense fingers (S) (shown as insets).
  • FIG. 3 is an optical microscope image of a microresonator with zoomed-in SEM views of a drive comb and a crab-leg spring.
  • FIG. 4 is a chart illustrating hysteresis of a Duffing resonator, indicating point A as the down-sweep bifurcation frequency and point B as a control point that is just below the up-sweep bifurcation frequency.
  • FIG. 5 is a bifurcation diagram of the Duffing resonator showing the locus of the up and down bifurcation points, and the states for the control, where one is located in the high displacement region and the other in the low displacement region.
  • FIG. 6 is the block diagram of a bi-state Duffing resonator control system according to one embodiment.
  • FIG. 7 depicts charts of the turn-on transient of the bi-state control loop for a Duffing resonator.
  • FIG. 8 depicts charts showing the recapture of the Duffing resonance by rapidly switching back to the “on” state.
  • FIG. 9 depicts charts illustrating the turn-off of the Duffing resonance by increasing the off-state time beyond the critical off-state time.
  • FIG. 10 is the zoom-in view of the frequency down sweep of a parametric resonator in the time domain, where the system state has to travel through the slow time manifold from “off” state (point A) to reach the “on” state (point B).
  • FIG. 11A illustrates the measured bifurcation diagram of the “on” (shaded) and “off” regions for a parametric resonator. Callout shows two servo states that are marked as circles.
  • FIG. 11B is a composite view that shows the relationship between the measured bifurcation of a parametric resonator and the corresponding frequency swept resonant amplitudes and that indicates the location of possible operating points at a select drive frequency.
  • FIG. 12 is a series of charts demonstrating the recapturing of the parametric resonance “zero-vicinity” state.
  • FIG. 13 illustrates the turn-on transient of the bi-state control loop for the parametric control embodiment.
  • FIG. 14 illustrates shifts in the servo setpoint for a parametric resonator illustrating closed-loop dynamics. Solid curves are obtained with only proportional control and the dash curves are obtained using proportional-derivative control.
  • FIG. 15 represents an Allan deviation plot of the temperature compensated controller duty cycle output quantifying the stability of a particular embodiment.
  • FIG. 16 is a flow diagram of the method of one embodiment of the present invention.
  • DETAILED DESCRIPTION OF THE INVENTION
  • A control system and method of the present invention makes possible the highly sensitive detection of mass or strain change by utilizing the bifurcation behavior inherent in a class of nonlinear parametric resonators 100 and in Duffing resonators 100. As shown in FIG. 1 and FIG. 4, the bifurcation behavior is represented as sharp increase in amplitude during a frequency sweep. Depending on the direction of travel of the frequency sweep, the bifurcation jump 501 occurs along the lower frequency transition or higher frequency transition.
  • FIG. 2 illustrates a block diagram of the bi-state bifurcation control system to serve as a general description of the resonator 100 and control loop 301 and is described in detail in the embodiments below. As shown in FIG. 2, the system is comprised of a parametric resonator 100 having capacitive drive combs 104 and capacitive sense combs 105. A direct current (DC) drive voltage is applied across the drive combs 104 to set the effective spring constant of the resonator 100. The sense combs 105 provide information used in a control loop 301 that switches the resonator 100 between “on” and “off” states. The combs 104 and 105 are comprised of a primary set of comb fingers 109 attached to the mass 102 and a corresponding set of fingers 108 that are mechanically anchored. The system further comprises a control loop 301, which may include a drive circuit 310. In one embodiment, the controller, or control loop, 301 further comprises a feedback circuit 308 and an error circuit 309.
  • As shown in FIG. 16, the method of control of the resonator 100 according to one embodiment is: at step 800, sensing a differential motional output current from the sense capacitors 105 of a MEMS parametric resonator 100; at step 801, generating a feedback signal; at step 802, extracting displacement and phase signal based on the resonator 100 output (in this embodiment using a lock-in amplifier 303 with a low-pass filter 307 (LPF) having a 2 Hz cutoff frequency); at step 803, subtracting the displacement signal, υx, with a pre-set threshold, Vthresh, and generating an error signal, εamp; at step 804, inputting the error signal into the proportional-integral-derivative controller; at step 805, generating the control signal with the proportional-integral-differential (PID) controller 306 based on the error signal; at step 806, inputting the control signal into the pulse-width-modulator (PWM) 302; at step 807, generating the pulse-width modulated bi-state control signal, υPWM, using the pulse-width modulation module 302; at step 808, using the bi-state control signal to modulate an AC drive signal, υac, wherein the AC drive signal is at twice the mechanical resonance frequency, ωr, of the resonator 100; at step 809, generating the electrical drive signal to the resonator 100 based on the summation of the pulse-width modulated bi-state control signal, υPWM, and the AC drive signal, υac; and at step 810, generating the electromechanical force signal to the resonator 100 proportional to the product of the bi-state control signal and the AC drive signal.
  • In one embodiment, a parametric control scheme is used to control a MEMS resonator 100 operating under nonlinear parametric resonance conditions. The parametric control scheme addresses the issue of the slow time manifold by servoing along this manifold with a bi-state PWM switching that is set to a speed much faster than the amplitude response dynamics of the resonator 100. In the embodiment shown in FIG. 2, the PWM switching speed is set to 200 kHz. However, other switching speeds can be used as long as the speed is faster than the resonator 100 system dynamics. The servo point 500 is chosen to take a non-zero displacement amplitude (1.9 μm, shown as point C in FIG. 10) that is slightly smaller than the maximum “on” amplitude (2.2 μm). The large displacement servo gives rise to a much higher ratio of signal to noise compared to the current state-of-the-art, which uses a servo point near zero amplitude (e.g., 100 pm). The servo point 500 lies along the bifurcation jump 501 A-B in FIG. 1. The servo point 500 can be set along the bifurcation jump point 501 S-P in FIG. 1, in an alternate embodiment.
  • In the preferred embodiment, the control plant 101 is a parametric resonator 100 made in a Silicon-On-Insulator (SOI)-MEMS fabrication process with a 15 μm-thick structural silicon device layer forming a micromechanical mass-spring system. Two primary sets of electrostatic comb fingers 109 are attached to the plate mass 102. Corresponding sets of comb fingers 108 are interdigitated with the primary set of comb fingers 109 attached to the mass 102 to form a drive comb 104 and a sense comb 105. The electrostatic sense fingers 109 which forms part of the sense comb 105 (a scanning electron microscope (SEM) image shown in FIG. 2, inset S) have straight sidewalls and act as capacitors whose capacitance varies linearly with displacement. The electrostatic drive comb fingers 109 which form part of the drive comb 104 (a SEM image shown in FIG. 2, inset D) is specially shaped along its sidewalls (for example, non-straight sidewalls) to create a voltage controlled linear spring constant γ1 in the force, Fp, acting on the resonator 100 in Eq. (1) that is necessary to excite the parametric resonance. A cubic-term spring constant γ3 in Eq. (1) provides specific nonlinear behavior that limits the parametric resonance amplitude. The specific effective system nonlinearity can be either hardening (i.e., γ3>0) or softening (i.e., γ3<0), which is determined by the combined mechanical and electrostatic nonlinearity. The exemplary embodiment of the control plant 102 in FIG. 2 has a hardening nonlinearity. The bi-state control principle applies to both hardening and softening nonlinear resonators.

  • F p(t,x)=−(γ1 x+γ 3 x 3)└|v dc|2 +|v ac|2 cos(ωe t)┘  (1).
  • In one example of a preferred embodiment, a DC bias voltage Vdc,set is applied to a total of 278 moving straight fingers 109 that are attached to a mass 102 of 9 μg. However, any number of fingers 108 and 109 can be used with sensitivity of the system increasing as the number of fingers 108 and 109 increases. As previously stated, the DC bias voltage sets the effective spring constant of the resonator 100. A voltage υshape formed by summing the DC voltage from the PWM 302, υPWM, with an AC excitation voltage amplitude, Vac, is applied on the shaped fingers 108 that are mechanically anchored. The other side of the comb 104, i.e. the fingers 109 attached to the moving plate mass 102, is connected to a fixed DC polarization voltage Vdc. Capacitive sensing circuits 107 use the set of linear capacitive comb fingers 105, with capacitance Cs, and a pair of transimpedance amplifiers 304 (TIA) to measure the differential motional current, ix=Vdc dCs/dt and subsequently the velocity amplitude, |x|, and phase of the moving mass 102.
  • Adjusting the total DC parametric polarization voltage that is applied on the resonator 100 Vdc,total=VdcPWM shifts the entire frequency response curve in FIG. 1 and as a result, shifts the bifurcation frequency. This effect is illustrated in FIG. 11 linking the bifurcation frequency to the DC parametric polarization voltage Vdc,total. With Vdc,total varying from 0 to 60 V, solid and dash lines correspond to the “pitch-fork” bifurcation frequencies at points A and C in FIG. 1, respectively. Increasing Vdc,total by adjusting the instantaneous value of υPWM, which is controlled by the feedback loop, shifts the frequency response curve (solid curve in FIG. 1) to the left and decreases the bifurcation jump 501 frequency. The specific correlation of the bifurcation frequency characteristic and the resulting frequency response curves is shown in FIG. 11B. The two PWM states shift the bifurcation points in the frequency response curve between “on” and “off” states when driven at the fixed AC drive frequency.
  • In both embodiments, the DC parametric polarization voltage Vdc,total takes two values, one value that drives the device into parametric resonance, and the other value drives the resonator to zero displacement. The bi-state control is accomplished by fixing the parametric drive frequency and the AC drive amplitude Vac, such that the system can be readily moved between “on” and “off” states with appropriate value of Vdc,total (circles in FIG. 11B). The constraint that the parametric polarization voltage is only allowed to take one of the two possible values (one in the “on” region, and the other in the “off” region) highlights a key characteristic of this control scheme. In contrast, a linear control scheme having a continuum of polarization voltage levels between the two state values would result in an unstable controller and would not servo on the desired setpoint. Instead, such a control system would latch into one of the “latched”, steady-state on (s.s. on in FIG. 11B) or steady-state off (s.s. off in FIG. 11B) positions and thus would not track the bifurcation as an average of the PWM 302 output.
  • To illustrate the control dynamics, a first DC parametric polarization voltage Vdc,total=49.7 V lands inside the “on” region, starting to move the state along the slow time manifold in FIG. 10 from point A to B. Before the state completely settles to point B, a second DC parametric polarization voltage Vdc,total=53.2 V is applied which resides in the “off” region, moving the state back from point B to A. This “recapture” dynamics is shown in FIG. 12. The successful recapturing of the resonator 100 state is guaranteed provided that the “on-state” time F is kept shorter than 602.8 ms, in this particular example. The recapturing breaks down if the on-state time is set beyond this critical time. Referring again to FIG. 12, recapturing occurs by rapidly switching back to the “off” state through the use of the control loop 301, with the pulse width modulator 302 alternating the DC voltage based on a set clock speed, as illustrated in FIG. 2.
  • In a practical control loop 301 that servos on large amplitude, when backing off from B to A, before reaching point A, Vdc,total is switched back to the “on” state again, inhibiting the “latch-off”, i.e., the displacement reduces to zero. The imposition of the bi-state parametric polarization voltages within the control loop 301 sets the servo point 500 at C in FIG. 10 with an appropriate switching speed (in the example device, an external signal source generates the 200 kHz clock signal) that is much faster than the slow sinusoidal-amplitude system dynamics (typically on the order of 0.5 Hz). Higher switching speeds are possible with appropriately fast electronics and slower switching speeds are possible with a trade-off of increased ripple in the servo amplitude. If the switching speed approaches the slow sinusoidal-amplitude system dynamics, the amplitude will latch states and not servo on the setpoint.
  • Referring again to FIG. 2, the physical implementation of the bi-state control scheme is schematically shown as a block diagram. A shaped-finger comb parametric resonator 100 serves as the control plant 101. A 3.3 V parametric AC drive voltage amplitude at 2ωr r=6910.7 Hz) is generated by an external frequency doubler circuit 110. With appropriate values of υdc, this twice resonance frequency component drives the oscillator into parametric resonance. The amplitude and the phase of the velocity signal υx, picked off by the TIA 304, are measured by a lock-in amplifier (LIA) 303 which utilizes the signal at ωr to demodulate the oscillation velocity signal at resonance. The velocity amplitude output (27.1 mV, as an example) is compared with a preset threshold Vthresh=25 mV to get the error signal εamp. This error signal is fed into a proportional-integral-derivative (PID) controller 306 that provides the duty cycle information to the pulse-width-modulation (PWM) module 302. The PWM 302 generates a square waveform with a fixed frequency at 200 kHz, outputting either high or low voltages that correspond to the first DC voltage or the second DC voltage. The PID 306 coefficients are chosen in such a way that the width ratio of the “on” and “off” states is changed to maintain the servo point between the maximum “on” and zero “off” states. This example shows the servo 500 at point C in FIG. 10 with 1.9 μm amplitude to yield a large ratio of servo amplitude to the minimum detectable amplitude. The overall parametric drive voltage across the drive comb fingers 104 is set by the PWM signal υPWM, the AC drive υac, and the DC bias voltage Vdc on the mass 102.

  • υparametric =|V dc−υacpwm|  (2).
  • The duty cycle signal is an ultra-sensitive indicator of the servo setpoint 500 along the slow sinusoidal-amplitude dynamics, therefore it is chosen as the controller output υo, derived by sending the PWM signal through a low pass filter (LPF) 307. The key characteristic of this design ensures that the system is servoing on the slow time manifold by moving between the zero “off” and non-zero “on” regions using the PWM fine-grained control.
  • Turn-on dynamics of this bi-state control loop in a vacuum of 18 mTorr, by way of one particular example, is shown in FIG. 13. At 12.8 s, a step change in the parametric drive frequency turns the resonator 100 to the “on” state (2.2 μm) from the “off” state. The controller 301 is turned on at 23.3 s by enabling the PID controller 306. The exemplary proportional and derivative controller 306 coefficients are kp=−141 V/V and kd=−7.1 V-s/V, respectively in this example. The servo point 500 (1.9 μm) backs off from the maximum “on” and settles on the slow time manifold. The duty cycle rings between 0.69 to 0.2 as the servo turns on and then settles to a steady-state control value of around 0.34.
  • Closed-loop dynamics shown in FIG. 14 illustrate the effect of the derivative controller coefficient on the closed-loop dynamics. The servo threshold, Vthresh, is changed at 0 s from 25 mV (equivalent to 1.75 μm in displacement amplitude |x|) to 28 mV (1.97 μm). With kp=−141 V/V and kd=0, the displacement amplitude changes from 1.89 μm to 2.04 μm with a significant amount of ringing (solid line). This ringing is alleviated by introducing a derivative control coefficient, kd=−7.1 V-s/V, which results in a near critically damped response (dashed line).
  • The bi-state control scheme for parametric resonators 100 is general and applies to any bi-state feedback scheme, of which the highlighted PWM scheme is a good example. Other possible feedback schemes include pulse-density modulation, for example in delta modulator or delta-sigma modulator architectures. Potentially, other discrete multi-state control schemes could also be employed instead of simply two states in feedback. An example would be in pulse amplitude modulation with a finite number of states. The preceding list is not meant to be exhaustive as many other digital and discrete-amplitude feedback methods exist.
  • In another embodiment, the system is a bi-state control system capable of servoing on the falling edge of the microelectromechanical Duffing resonant bifurcation setpoint. As shown in FIG. 3, the system is presented with a 15 μm-thick SOI-MEMS Duffing resonator 100 as the control plant 101, which is comprised of a shuttle mass 102 and a set of capacitive comb fingers 109. Further shown in FIG. 3 is a perspective view of the lateral capacitive combs 104 and 105 electrode geometry of the control plant 101. The capacitive combs 104 and 105 are symmetrically placed on both sides of the shuttle mass 102; the left set serves as the drive electrode 104 and the right set serves as the capacitive pick-off sense electrode 105. The perforated shuttle mass 102 is suspended by four symmetric crab-leg springs, or spring suspension, 106. The nonlinearity inherent in the system stems from the suspension and the nonlinear electrostatic spring constants arising from the comb drive geometry.
  • The typical measured frequency response of a Duffing resonator 100 is plotted in FIG. 4. Referring to FIG. 4, ↑ symbolizes the frequency up-sweep and ↓ is the frequency down-sweep. This class of nonlinear resonator 100 exhibits a hysteresis effect when the frequency is swept bi-directionally. In the example depicted in FIG. 4, the voltage on the moving shuttle mass 102 is Vdc,set=10 V and is called the “polarization” voltage. The magnitude of the AC sinusoidal drive on the lateral capacitive comb rotor is Vac=25 mV. In this embodiment, the numbers depicted in FIG. 4 correspond to a test performed in a low-pressure environment of 25 mTorr to serve as an illustrative example.
  • Unlike the linear harmonic resonator, the resonance peak bends toward higher frequencies, yielding multi-valued solutions for any particular frequency value located around the resonance. For example, on the edge of the bent peak at the maximum amplitude (point B on FIG. 4), a jump in the resonance amplitude from 4.8 μm to 0 μm occurs if the drive frequency is increased an infinitesimal amount. The frequency at which the instantaneous jump occurs is called the “bifurcation frequency” fB, which equals 4167 Hz for the setpoint B in FIG. 4. However, the bifurcation frequency can vary depending on the design of the resonator. For a reverse frequency sweep, the jump 501 happens at a lower frequency from 0 μm to 1.0 μm (point A), indicative of a hysteresis effect. The region I between points A and B defines the region where a high amplitude response occurs in response to a frequency change. To the right of point B (region II), the resonance amplitude reduces to very low displacement.
  • The states for a bi-state control loop can be defined by adjusting the AC drive voltage Vac which shifts the entire frequency response curve in FIG. 4, thus resulting in changes in the bifurcation points A and B. These shifts in A and B are graphically represented in FIG. 5, which defines the “instability tongue” with the locus of the amplitude jump points. The right curve is the up-sweep bifurcation frequency (analogous to point B in FIG. 4) showing the falling edge of the instability; the left curve is the down-sweep bifurcation frequency (point A). The DC polarization voltage is fixed at 10 V in the example loci in FIG. 5.
  • The changes in the drive voltages modify the system's effective spring constant, which affects the bending of the resonance curve and the regions where the system enters instability. In a typical nonlinear resonator, the drive voltage shapes the instability tongue. The tongue width measured in units of frequency (e.g., Hertz) depends on the AC drive amplitude. The position of the tongue along the frequency axis is a strong function of the DC polarization voltage.
  • The points 502 in FIG. 5 identify the states for a bi-state control loop. One of the states (“on” state) is inside the “tongue” that has a large amplitude response; the other state (“off” state) resides outside the “tongue” representing a low amplitude response. The two states center around Vac=25 mV and have the same fixed frequency at the bifurcation frequency fB for the Vac=25 mV setpoint. The two states are on opposite sides of the bifurcation point in steady-state operation. Adjusting the AC voltage, Vac, near the boundary can move the control state between high and low displacement values. The rapid switching between these two states determines the servo setpoint.
  • The control loop 301 is schematically shown in FIG. 6. The control setpoint 500 of the loop 301 is chosen at point B in FIG. 4. The AC drive frequency and the DC bias are fixed to values such that the system can be moved between high and low displacement regions with appropriate bi-state values of AC amplitude |Vac|. In one particular example, the DC polarization voltage Vdc,set=10 V and the AC drive frequency is fB=4167 Hz. The Duffing resonator 100 is driven by Vac whose amplitude and frequency are set by the pulse-width-modulator (PWM) 302 output and the lock-in amplifier (LIA) 303, respectively. A transimpedance amplifier 304 followed by the LIA 303 picks up the resonator velocity amplitude |v{tilde over (x)}|. The loop 301 tracks |v{tilde over (x)}| to a fixed setpoint (Vthresh). The error output is fed into a PID (proportional-integral-derivative) controller 306 to generate the input signal to the PWM 302 representing the pulse-train duty cycle. The output of the PWM 302 generates the bi-state control signal that sets the AC drive amplitude.
  • The Vac is controlled by the duty-cycle feedback and rapidly adjusts the oscillation between the two states (e.g., high and low displacements) at an example PWM rate of 300 kHz to achieve the servo at the maximum displacement value without turning off the resonator 100. This avoids the long ring-down transient needed to settle from the high displacement state to the low displacement state. This resettling time is inversely proportional to the damping and is typically on the order of 1 s for the resonator 100 operated under a similar vacuum level.
  • Higher PWM 302 switching rates are possible with appropriately fast electronics and slower switching speeds are possible with a trade-off of increased ripple in the servo amplitude. If the switching speed approaches the slow amplitude Duffing system dynamics (proportional to 2Q/ωr, where Q is the quality factor and ωr is the resonant frequency), the amplitude will latch to the zero state and not servo on the high displacement setpoint.
  • FIG. 7 shows a series of graphs which depict the turn-on transient of the closed-loop controller 301 with an initial approaching phase. In FIG. 7, the drive AC voltage is turned on at 15 s, followed by an approach phase when the frequency is increased from 4123 Hz to 4167 Hz. The controller 301 is turned on afterwards. At 75 s, the controller 301 is turned off by switching off Vac. During region (1), the resonator is off. During region (2), a sinusoidal voltage at 4123 Hz is applied to the resonator 100 drive fingers of comb 104, resulting in a step turn on of the displacement amplitude to a non-zero value due to the capacitive feedthrough. The choice of the initial frequency ensures the resonator 100 is within its low displacement state (to the left of point A in FIG. 4). Region (3) is the necessary “approach” stage when the resonator 100 is turned on to the maximum displacement operation point (point B in FIG. 4) by increasing the drive frequency from 4123 Hz to 4167 Hz. The PID controller 306 threshold Vthresh 305 is set in region (4) after reaching this maximum operation point, turning on the servo, after which time the PWM 302 output rapidly changes between 0.3 V and 3.4 V, with duty cycle centered on 0.5. At region (5), the servo is switched off by turning off Vac. The phase is well-defined during the time the controller 301 is on and loses track of the input Vac when the controller 301 is turned off. A detailed analysis of the dynamics on the edge of the bifurcation jump 501 provides information on how fast the controller 301 should switch between two states to guarantee successful recapturing of the “on” state. These “off-state” dynamics are characterized in FIG. 8 and FIG. 9.
  • In open-loop operation, a step change in Vac is introduced from 25.6 mV to 24.1 mV, at which voltage the resonator 100 is not fully in the off state but rather just starting to turn off. Vac is then turned back to 25.6 mV after a prescribed amount of time Γ. When Γ is less than 394.6 ms, the Duffing resonance is successfully recaptured as presented in FIG. 8. When Γ is larger than 395.0 ms, the resonance falls back to the low displacement state. The critical off-state time Γ at which the resonance is on the edge of being recaptured is 394.6 ms.
  • Similar to the “recapture” dynamics in FIG. 12, in both FIGS. 8-9 prior to t=0 (point F), the system operates at its maximum “on” state. After point F, the resonator 100 starts to turn off and provides the initial condition for later step turn-on (point N). At point N, the system exhibits a step response due to the step turn-on in the drive amplitude Vac. Depending on the length of the “off-state” time (the time between points F and N), the system can either recover from the decreased amplitude (FIG. 8) or completely turns off (FIG. 9). The bi-state control signal must operate at a rate much higher than the critical off-state time to avoid the ringing. The recapture dynamics in FIG. 8 gauge the “slow time” (i.e., time scale that is much longer than the resonance period) behavior of the Duffing resonator 100 and thus provide the root-locus information for later optimization of the PID 306 coefficients. In FIG. 8, the critical off-state time is 394.6 ms. In FIG. 9, the turn-off time is 395.0 ms.
  • Allan standard deviation of the controller 301 duty cycle output for a parametric resonator 100 is shown in FIG. 15 that represents the servo point 500 variation versus the integration time, τ, and demonstrates the stability of the controller 301. The duty cycle signal is temperature compensated to eliminate the effect of temperature on the bifurcation frequency variation and hence the servo setpoint 500. As an example, the data shown in FIG. 15 represent a test performed in a constant vacuum of 18 mTorr. In this particular example, the minimum detectable duty cycle found at an integration time of 4.3 s is 2.36×10−4, which corresponds to 4.48 Å resolution in displacement amplitude |x|. The low-pass filter 307 attenuates high frequency noise beyond a bandwidth of 1.95 Hz above and below the mechanical resonance at ωr=6910.7 Hz. Calibrating the gain from the duty cycle to the minimum detectable mass variation enables use of the system for a sensitive gravimetric sensor. The large displacement (1.9 μm) operation yields a much higher ratio of operation amplitude to the minimum detectable signal compared with recent bifurcation-based control schemes.
  • The embodiments and examples described herein demonstrate a nonlinear control of parametric and Duffing resonators 100 using a bi-state switching methodology, and further builds an ultra-sensitive platform for detecting parametric parameter changes that are indicative of resonance frequency shifts. The advantage of the system and method of the present invention is its design allowing for on-chip or off-chip implementation, and its ability to servo at relatively large amplitude (roughly four orders of magnitude larger in the displacement amplitude) along the sharp jump of the bifurcation without altering the plant. This bi-state closed-loop control system can be used in any application where mechanical resonant sensors are used now. Other types of nonlinear MEMS, nanomechanical, piezoelectric, and electrothermal devices that operate in resonance and experience bifurcation phenomena would benefit from the discrete-state control approach. The method described herein is adaptable for building robust sensor technologies including but not limited to, bifurcation-based gravimetric (mass) or chemical sensors. The most common application of the latter is in humidity sensing, widely used and commercialized on the market. This invention can also lead to the monolithic-chip implementation of the control electronics with the on-chip MEMS parametric resonator 100 to make a single-chip sensor suite.
  • While the system and method have been described in detail and with reference to specific embodiments and examples thereof, it will be apparent to one skilled in the art that various changes and modifications can be made therein without departing from the spirit and scope of the embodiments. Thus, it is intended that the present disclosure cover the modifications and variations of this disclosure provided they come within the scope of the appended claims and their equivalents.

Claims (19)

What is claimed is:
1. A system for controlling a resonator at a bifurcation frequency, comprising:
a resonator;
a control loop coupled to the resonator, comprising:
a feedback circuit for generating a feedback signal based on the output voltage of the resonator;
an error circuit coupled to the feedback circuit, wherein the error circuit generates an error signal based on a differential between the feedback signal and a threshold value;
a proportional-integral-derivative controller coupled to the error circuit, wherein the proportional-integral-derivative controller generates a control signal based on the error signal;
a pulse-width modulator coupled to the proportional-integral-derivative controller, wherein the pulse-width modulator generates a bi-state control signal based on the control signal and a preset clock speed;
a drive circuit coupled to the control loop and the resonator, wherein the drive circuit receives the bi-state control signal, wherein the drive circuit generates a drive signal for the resonator based on the bi-state control signal and a reference value.
2. The system of claim 1, wherein the resonator is a parametric resonator.
3. The system of claim 1, wherein the resonator is a Duffing resonator.
4. The system of claim 2, wherein the reference value is based on a resonance frequency of the resonator.
5. A system for switching a resonator between states at a bifurcation frequency, comprising:
a resonator comprising:
a mass;
a spring suspension;
a first set of capacitive comb fingers interdigitated with a first set of corresponding comb fingers on the mass;
a second set of capacitive comb fingers interdigitated with a second set of corresponding comb fingers on the mass;
wherein the first set of capacitive comb fingers drives an effective spring constant of the resonator based on a drive voltage applied across the first set of capacitive comb fingers and the first set of corresponding comb fingers;
wherein the second set of capacitive comb fingers produces an output signal; and
a control loop that generates a bi-state control signal, wherein the bi-state control signal switches the resonator between states in relation to a servo amplitude point at a bifurcation frequency.
6. The system of claim 5, wherein the control loop comprises:
a transimpedance amplifier for receiving the output signal and generating a modified output signal;
a lock-in amplifier, wherein the lock-in amplifier processes the modified output signal to determine the resonator velocity amplitude;
an error circuit that compares the velocity amplitude to a threshold value, generating an error signal;
a proportional-integral-derivative controller which receives the error signal and creates an adjusted drive signal;
a pulse width modulator that receives the adjusted drive signal and generates the bi-state control signal.
7. The system of claim 5, wherein the bi-state control signal is a square wave-form signal alternating between a first voltage and a second voltage.
8. The system of claim 5, wherein the drive voltage is a DC voltage.
9. The system of claim 5, wherein the drive voltage is an AC voltage.
10. The system of claim 5, wherein the effective spring constant is a linear constant.
11. The system of claim 5, wherein the effective spring constant is a nonlinear constant.
12. The system of claim 5, wherein the bi-state control signal is a DC voltage, wherein the DC voltage is impressed across the first set of capacitive comb fingers.
13. The system of claim 12, wherein the DC voltage is at least one discrete value.
14. The system of claim 5, wherein the bi-state control signal is comprised of two discrete DC voltage values.
15. The system of claim 5, further comprising a parametric drive circuit, wherein the parametric drive circuit impresses an AC voltage across the first set of capacitive comb fingers.
16. The system of claim 15, wherein the AC voltage has a fixed frequency.
17. A method of controlling a resonator at a bifurcation frequency having a control loop comprised of a proportional-integral-differential controller and a pulse-width modulator, the method comprising:
sensing a differential output current from the resonator;
generating a feedback signal;
extracting a displacement signal and a phase signal based on the differential output current;
subtracting the displacement signal and the phase signal with a preset threshold value to create an error signal;
inputting the error signal into the proportional-integral-differential controller;
generating a control signal with the proportional-integral-differential controller based on the error signal;
inputting the control signal into the pulse-width modulator;
generating a bi-state control signal;
using the bi-state control signal to modulate an ac drive signal, wherein the ac drive signal is integer multiple of the resonance frequency of the resonator;
generating the electrical drive signal to the resonator based on the summation of the bi-state control signal and the ac drive signal; and
generating the electromechanical force signal to the resonator proportional to the product of the bi-state control signal and the ac drive signal.
18. The method of claim 17, wherein generating a feedback signal further comprises:
combining the differential output current with a generated signal, wherein the generated signal has a frequency equal to an integer multiple of a resonance frequency of the resonator.
19. The method of claim 17, wherein the control signal includes duty cycle information.
US14/627,801 2014-02-20 2015-02-20 Method and device for bi-state control of nonlinear resonators Active 2035-03-30 US10608614B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US14/627,801 US10608614B2 (en) 2014-02-20 2015-02-20 Method and device for bi-state control of nonlinear resonators

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US201461966280P 2014-02-20 2014-02-20
US14/627,801 US10608614B2 (en) 2014-02-20 2015-02-20 Method and device for bi-state control of nonlinear resonators

Publications (2)

Publication Number Publication Date
US20160072472A1 true US20160072472A1 (en) 2016-03-10
US10608614B2 US10608614B2 (en) 2020-03-31

Family

ID=55438474

Family Applications (1)

Application Number Title Priority Date Filing Date
US14/627,801 Active 2035-03-30 US10608614B2 (en) 2014-02-20 2015-02-20 Method and device for bi-state control of nonlinear resonators

Country Status (1)

Country Link
US (1) US10608614B2 (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160336810A1 (en) * 2015-05-12 2016-11-17 The Regents Of The University Of Michigan Nonlinear Resonance Circuit For Wireless Power Transmission And Wireless Power Harvesting
US10784723B2 (en) 2015-05-12 2020-09-22 The Regents Of The University Of Michigan Nonlinear resonance circuit for wireless power transmission and wireless power harvesting
US20210356272A1 (en) * 2020-05-13 2021-11-18 Robert Bosch Gmbh Method for operating a microelectromechanical gyroscope, and gyroscope
WO2021229941A1 (en) * 2020-05-15 2021-11-18 パナソニックIpマネジメント株式会社 Resonant sensor using mems resonator and detection method for resonant sensor
US11513135B2 (en) * 2018-11-13 2022-11-29 Albert-Ludwigs-Universität Freiburg Method for automatic frequency adaptation of a filter in a closed loop

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5442510A (en) * 1993-06-23 1995-08-15 The United States Of America As Represented By The Secretary Of The Navy Control system for tracking nonlinear systems
US6497141B1 (en) * 1999-06-07 2002-12-24 Cornell Research Foundation Inc. Parametric resonance in microelectromechanical structures
US20090309613A1 (en) * 2008-06-11 2009-12-17 Analog Devices, Inc. Anti-Capture Method and Apparatus for Micromachined Devices
US20130009716A1 (en) * 2011-07-06 2013-01-10 Nxp B.V. Mems resonator
US8384372B1 (en) * 2005-02-14 2013-02-26 Clemson University Non-linear electrical actuation and detection
US20140132186A1 (en) * 2012-11-14 2014-05-15 Daniel A. Tazartes Amplitude control for vibrating resonant sensors

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5442510A (en) * 1993-06-23 1995-08-15 The United States Of America As Represented By The Secretary Of The Navy Control system for tracking nonlinear systems
US6497141B1 (en) * 1999-06-07 2002-12-24 Cornell Research Foundation Inc. Parametric resonance in microelectromechanical structures
US8384372B1 (en) * 2005-02-14 2013-02-26 Clemson University Non-linear electrical actuation and detection
US20090309613A1 (en) * 2008-06-11 2009-12-17 Analog Devices, Inc. Anti-Capture Method and Apparatus for Micromachined Devices
US20130009716A1 (en) * 2011-07-06 2013-01-10 Nxp B.V. Mems resonator
US20140132186A1 (en) * 2012-11-14 2014-05-15 Daniel A. Tazartes Amplitude control for vibrating resonant sensors

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160336810A1 (en) * 2015-05-12 2016-11-17 The Regents Of The University Of Michigan Nonlinear Resonance Circuit For Wireless Power Transmission And Wireless Power Harvesting
US10199869B2 (en) * 2015-05-12 2019-02-05 The Regents Of The University Of Michigan Nonlinear resonance circuit for wireless power transmission and wireless power harvesting
US10784723B2 (en) 2015-05-12 2020-09-22 The Regents Of The University Of Michigan Nonlinear resonance circuit for wireless power transmission and wireless power harvesting
US11513135B2 (en) * 2018-11-13 2022-11-29 Albert-Ludwigs-Universität Freiburg Method for automatic frequency adaptation of a filter in a closed loop
US20210356272A1 (en) * 2020-05-13 2021-11-18 Robert Bosch Gmbh Method for operating a microelectromechanical gyroscope, and gyroscope
WO2021229941A1 (en) * 2020-05-15 2021-11-18 パナソニックIpマネジメント株式会社 Resonant sensor using mems resonator and detection method for resonant sensor

Also Published As

Publication number Publication date
US10608614B2 (en) 2020-03-31

Similar Documents

Publication Publication Date Title
US10608614B2 (en) Method and device for bi-state control of nonlinear resonators
US9989553B2 (en) Extracting inertial information from nonlinear periodic signals
US8875578B2 (en) Electronic damper circuit for MEMS sensors and resonators
US9709400B2 (en) System, apparatus, and method for resonator and coriolis axis control in vibratory gyroscopes
FI126019B (en) MEMS resonator drive current circuitry
FI125611B (en) Traction circuit for activating the MEMS resonator
Agrawal et al. Modeling nonlinearities in MEMS oscillators
TWI596885B (en) A control element, a resonator device, and a method for controlling operation of a mechanical resonator
US8773004B2 (en) Circuit for optimizing the recovery of vibratory energy by a mechanical/electrical converter
WO2010011288A1 (en) Switching devices and related methods
AU2012344309B2 (en) Regulating device, rotation rate sensor and method for operating a regulating device with harmonic desired value signal
CN102753936B (en) Oscillation type inertial force sensor
CN106885563B (en) Micro-mechanical gyroscope closed-loop driving circuit capable of preventing electric oscillation
EP3146294B1 (en) Stepped sinusoidal drive for vibratory gyroscopes
RU2565516C2 (en) Control module and device for return to initial state of oscillator excited by harmonic oscillation, as well as sensor of instantaneous angular speed
US9611139B2 (en) Resonator
US11320265B2 (en) Offset-cancelling capacitive MEMS gyroscope
JP2008033567A (en) Piezoelectric element control method, piezoelectric element control device, actuator, and microscope
Van Leeuwen et al. Nonlinear dynamics of a microelectromechanical oscillator with delayed feedback
Guo et al. Bi-state control of a duffing microresonator on the falling edge of the instability
WO2016187560A1 (en) Extracting inertial information from nonlinear periodic signals
Guo et al. Bi-state control of parametric resonance
Taheri-Tehrani et al. Epitaxially-encapsulated quad mass resonator with shaped comb fingers for frequency tuning
Czaplewski et al. Nonlinear mode coupling in a MEMS resonator
Bouchami et al. Non-linear modeling of MEMS-based oscillators using an analog hardware description language

Legal Events

Date Code Title Description
AS Assignment

Owner name: CARNEGIE MELLON UNIVERSITY, A PENNSYLVANIA NON-PROFIT CORPORATION, PENNSYLVANIA

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:FEDDER, GARY K.;GUO, CONGZHONG;SIGNING DATES FROM 20150519 TO 20150724;REEL/FRAME:036181/0843

Owner name: CARNEGIE MELLON UNIVERSITY, A PENNSYLVANIA NON-PRO

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:FEDDER, GARY K.;GUO, CONGZHONG;SIGNING DATES FROM 20150519 TO 20150724;REEL/FRAME:036181/0843

AS Assignment

Owner name: NATIONAL SCIENCE FOUNDATION, VIRGINIA

Free format text: CONFIRMATORY LICENSE;ASSIGNOR:CARNEGIE-MELLON UNIVERSITY;REEL/FRAME:038269/0384

Effective date: 20150510

STPP Information on status: patent application and granting procedure in general

Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION

STPP Information on status: patent application and granting procedure in general

Free format text: NON FINAL ACTION MAILED

STPP Information on status: patent application and granting procedure in general

Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER

STPP Information on status: patent application and granting procedure in general

Free format text: NOTICE OF ALLOWANCE MAILED -- APPLICATION RECEIVED IN OFFICE OF PUBLICATIONS

STPP Information on status: patent application and granting procedure in general

Free format text: PUBLICATIONS -- ISSUE FEE PAYMENT VERIFIED

STCF Information on status: patent grant

Free format text: PATENTED CASE

MAFP Maintenance fee payment

Free format text: PAYMENT OF MAINTENANCE FEE, 4TH YR, SMALL ENTITY (ORIGINAL EVENT CODE: M2551); ENTITY STATUS OF PATENT OWNER: SMALL ENTITY

Year of fee payment: 4